JP2004279186A - Method and apparatus for estimating light source energy distribution, and method for determining exposure amount - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To estimate the spectral energy distribution of a photographic light source from an image signal recorded under a prescribed photographing condition, and to apply the result for controlling exposure light. <P>SOLUTION: The combination of main component vectors, which minimize the difference between the sensor output value data of the image recorded with a light having an unknown spectral energy distribution and sensor output value data corresponding to the spectral energy distribution formed by the main component vectors consisting of a variety of combinations, are made as the spectral energy distribution of the photographic light source. From the transmitted concentration of the optical measurement of a negative film (step 106), the evaluation value is obtained using the main component vectors consisting of the variety of combinations (step 108). It is estimated that the combination of the main component vectors providing the minimum obtained evaluation value is the spectral energy distribution of the photographic light source (step 110). <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

【００２１】
しかしながら、被写体の分光反射率分布は、撮影時の撮影画像に応じて異なるので、特定することができず、フィルム上の画像から撮影光源の分光エネルギー分布等を推定することが困難であったことは、前述の通りである。
【００２２】
これに対して、本発明に係る光源の分光エネルギー分布推定方法においては、未知の分光エネルギー分布を有する光源の分光エネルギー分布を、標準光源といわれる既知の各種の光源の主成分分析結果に基づく主成分ベクトルを求め、この主成分ベクトルをｖ_ｊ （λ）とするとき、以下の式（２）に従って光源の分光スペクトル分布を発生させるようにしたものである。
【００２３】
【数２】

【００２４】
ここでは、ｋ個の加重係数ｂ_ｎ をさまざまに変化させて、上記式（２）に従って光源の分光エネルギー分布Ｐ（λ）を発生させる。こうして発生させた光源の分光エネルギー分布Ｐ（λ）について、前述の特許文献１に開示されている方法で、評価値Ｖを求める。求められた評価値Ｖについて、最も小さいＶ値を与える加重係数ｂ_ｎ を求める。
【００２５】
そして、最小のＶ値を与える加重係数ｂ_ｎ について、上記式（２）に従って光源の分光エネルギー分布を計算し、得られた光源分光分布が、撮影に使用された光源の分光分布であると推測する。
以上が、本発明の作用の概要である。より詳細には、以下の通りである。
【００２６】
【発明の実施の形態】
未知の分光エネルギー分布を有する光源で照明された光源からの光で照明された被写体を撮影した場合のセンサ出力値データ（Ｅ_ｉｊ ^０ ）がある場合、その被写体が記録されたときの光源の分光エネルギー分布を求めることを想定すると、次の式（３）に示すように、データＥ_ｉｊ ^０ と、前述の、加重係数ｂ_ｎ をさまざまに変化させて、前述の式（２）に従って発生させた光源の分光エネルギー分布Ｐ（λ）に対応するセンサ出力値Ｅ_ｉｊ ^Ｔ との差ΔＥを最小にする分光エネルギー分布を有する光源が解である。
【００２７】
【数３】

【００２８】
上述の式（３）のΔＥの最小化演算を実行するため、ここでは、モデル演算として、ＪＩＳＺ８７２０の標準の光Ａ，Ｄ６５，Ｃおよび補助標準の光Ｄ５０，Ｄ５５，Ｄ７５の６種の光源と、ＪＩＳＺ８７１９に記載の１２種類の蛍光灯との、計１８種の光源分光スペクトルに対して、主成分分析を適用して、その主成分ベクトルを求めた。
【００２９】
図１〜図５に、その特性を示した。なお、ここで、図１は第１主成分を、図２は第２主成分をというように、第５主成分までを示している。
こうして求めた主成分ベクトルについて、前述の式（２）に従って種々の光源の分光スペクトル分布を発生させる。
【００３０】
ここでは、上述のｋ個の加重係数ｂ_ｊ をさまざまに変化させて、前記式（２）に従って多数の光源分光分布Ｐ（λ）を発生させる。こうして発生させた光源の分光エネルギー分布Ｐ（λ）について、被写体としてマクベスチャート２４色を用いた場合について、上記式（３）のΔＥの最小化を行った。なお、ここで、データＥ_ｉｊ ^０ は、実際のフィルムの分光感度および実測の分光反射率を用いて全記式（１）を変形した下記の式（４）から求めている。
【００３１】
【数４】

【００３２】
ところで、種々のｂ_ｎ の値に対応する光源の分光エネルギー分布Ｐ（λ）の中には、これに対応するものとして復元される被写体の分光反射率に異常な値を示すものが含まれる場合がある。すなわち、得られる分光反射率ρ_ｉ （λ）が負の値になったり、１．０を大幅に超える場合がある。
【００３３】
分光反射率ρ_ｉ （λ）は、負の値や１．０を超えることはない（つまり、０≦ρ_ｉ （λ）≦１）ので、逆に、この異常な分光反射率ρ_ｉ （λ）を示す元になった光源の分光エネルギー分布Ｐ（λ）が、「真の」光源の分光エネルギー分布Ｐ（λ）からずれていると想定することができる。
【００３４】
このため、下記の式（５）で示される評価値Ｖを導入する。
【数５】

【００３５】
そして、この評価値Ｖを、種々のｂ_ｎ の値に対応する光源の分光エネルギー分布Ｐ（λ）に対して演算し、この評価値Ｖが最小となるｂ_ｎ の組み合わせに対応する光源の分光エネルギー分布Ｐ（λ）を、実際の撮影に使用された光源の分光エネルギー分布Ｐ（λ）と推定する。
【００３６】
図６〜図９に、上述の方法により求めた、光源の分光エネルギー分布と、対象としたフィルム上の画像が実際に撮影されたときに用いられた光源の、いわば、「真の」分光エネルギー分布とを対比させて示す。
なお、ここでは、前述の主成分ベクトルを第３主成分まで用いた場合と、第５主成分まで用いた場合とも比較して示す。
【００３７】
まず、図６は、一般的なキセノンランプの分光エネルギー分布（実線で示されている）と、これを用いて撮影されたフィルム上の画像について、本発明に係る推定方法を適用して求めた「推定された」光源の分光エネルギー分布（一点鎖線で示されている）とを比較する図である。なお、ここでは、主成分ベクトルを第３主成分まで用いた場合を示している。
【００３８】
また、図８は、上述の組み合わせを、主成分ベクトルを第５主成分まで用いた場合を示している。両者を比較すれば明らかなように、主成分ベクトルを第３主成分まで用いた場合 （図６）に比べて、主成分ベクトルを第５主成分まで用いた場合（図８）には、実際の特性と、推定した結果との一致度に若干の差異が認められるが、主成分ベクトルを第３主成分まで用いた場合でも、実用的には、十分利用可能なレベルである。
【００３９】
次に、図７および図９は、市販の蛍光灯とタングステン電球とを両方用いた状況での撮影の場合を示すもので、両図において、実線で示されているのは、これらの２つの光源を両方点灯した状態における、組み合わせ状態の分光エネルギー分布、また、一点鎖線で示されているのは、この状況で、本発明に係る推定方法を適用して求めた「推定された」光源の分光エネルギー分布である。
【００４０】
なお、図７は、主成分ベクトルを第３主成分まで用いた場合を示しており、図９は、主成分ベクトルを第５主成分まで用いた場合を示している。この場合においても、主成分ベクトルを第３主成分まで用いた場合（図７）に比べて、主成分ベクトルを第５主成分まで用いた場合（図９）には、実際の特性と、推定した結果との一致度に若干の差異が認められるが、主成分ベクトルを第３主成分まで用いた場合でも、実用的には、十分利用可能なレベルである。
【００４１】
次に、参考までに、図７および図９に示した状況下で、主成分ベクトルを第２主成分まで用いて、本発明に係る方法による光源の分光エネルギー分布の推定を行った場合の結果を、図１０に示す。
図１０から明らかなように、主成分ベクトルを第２主成分まで用いる推定方法では、推定の精度が不十分であり、実用になるとはいいがたい。
【００４２】
以上の結果から、本発明に係る光源の分光エネルギー分布の推定方法においては、少なくとも主成分ベクトルを第３主成分まで用いて推定を行うことが好ましく、より好ましくは、主成分ベクトルを第５主成分まで用いて推定を行うことであるということができる。
【００４３】
前述の通り、本発明に係る光源の分光エネルギー分布の推定方法によって得られた光源の分光エネルギー分布の情報は、これを、写真プリント（焼き付け）時におけるプリントの露光時間の制御に応用することが可能である。
【００４４】
なお、フィルムの焼き付け対象画像を複写感材としてのプリント感材に焼き付ける際に、この推定した分光エネルギー分布を有する光源に基づいて、プリントの露光時間を制御するという応用を想定した場合には、前述のセンサの分光感度分布Ｓ_ｊ （λ）を、フィルムの感度分布、例えば図１１に示すようなカラーネガフィルムの感度分布に置き換えることで適用可能である。
【００４５】
ここで、フィルムから発色濃度を測光するための光源の分光エネルギー分布，プリント感材の分光感度分布は、予め測定することによって求められるから、フィルムによって異なるフィルムの分光透過率分布が定まれば、プリント感材の分光感度分布と等しい分光感度分布の測光装置で測光した濃度も求めることができる。この技術に関しては、特開平４−３１０９４２号公報に記載のフィルムの分光分布推定方法にも記載がある。
【００４６】
【実施例】
以下、具体的な実施例を上げて説明する。
本実施例は、本発明を、自動プリンタに適用したものである。
【００４７】
図１２は、上述の、推定した光源の分光エネルギー分布に基づく焼き付け方法によって写真を焼き付ける自動プリンタの概略図を示したものである。
ネガキャリア２１に装填されて焼き付け部に搬送されたカラーネガフィルム２０の下方には、ミラーボックス１８およびハロゲンランプを備えたランプハウス１０が配列されている。ミラーボックス１８とランプハウス１０との間には、調光フィルタ６０が配置されている。調光フィルタ６０は、周知のようにＹ（イエロー）フィルタ，Ｍ（マゼンタ）フィルタおよびＣ（シアン）フィルタの３つのフィルタで構成されている。
【００４８】
ネガフィルム２０の上方には、レンズ２２，ブラックシャッタ２４およびカラーペーパ２６が順に配置されており、ランプハウス１０から照射されて調光フィルタ６０，ミラーボックス１８およびカラーネガフィルム２０を透過した光が、レンズ２２によってカラーペーパ２６上に結像するように構成されている。
【００４９】
カラーネガフィルム２０の側縁部には、カラーネガフィルムの種類を表わすＤＸコートが記録されるとともに、ノッチが穿設されている。このＤＸコードやノッチを検出するために、ネガフィルム２０の側縁を挟むように、発光素子と受光素子とで構成された検出器５２が配置されている。
【００５０】
上に説明した結像光学系の光軸に対して傾斜した方向で、かつカラーネガフィルム２０の画像濃度を測光可能な位置に、測光器２８が配置されている。この測光器２８は、中心波長が４５０±５ｎｍ，５５０±５ｎｍ，７００±５ｎｍで半値幅が各々１５〜５０ｎｍの３つのフィルタと２次元イメージセンサとで構成されている。この測光器２８によって、カラーネガフィルムから透過した光を３つの波長帯に分光して測定することができる。
【００５１】
測光器２８は、測光器２８で測光された画像データを記憶する画像データメモリ３０を介してマイクロコンピュータで構成された露光量決定装置３２に接続されている。露光量決定装置３２は、入出力ポート３４，中央処理装置（ＣＰＵ）３６，リードオンリメモリ（ＲＯＭ）３８，ランダムアクセスメモリ（ＲＡＭ）４０およびこれらを接続するデータバスやコントロールバス等で構成されたバス４２を備えている。
【００５２】
このＲＯＭ３８には、以下で説明する露光量制御ルーチンのプログラムや図１３に示すカラーネガフィルムの固有ベクトルｅ_１ （λ）、ｅ_２ （λ）、ｅ_３ （λ）の分布がフィルム種毎に記憶されている。なお、図１３では、１種のフィルムに対する固有ベクトルの分布を示したが、他の種類のフィルムについても略同様である。また、このＲＯＭ３８には、ランプハウス１０内のハロゲンランプの分光エネルギー分布，使用するペーパーの分光感度分布，上記３つのフィルタの透過波長域に対応する測光器の３つの分光感度分布が予め記憶されている。
【００５３】
なお、ペーパーを変更する場合には、予めＲＯＭに複数種のペーパーの分光感度分布を記憶しておいてキーボードによって使用するペーパーの分光感度分布を選択してもよく、また、フレキシブルディスク等の外部メモリに記憶した必要なペーパーの分光感度分布を、ＲＡＭに読み込むようにしてもよい。また、ランプを交換したときには、ランプの分光エネルギー分布を変更するようにすることも好ましい。
【００５４】
そのために、ランプの分光エネルギー分布を直接またはフィルタを通して測光器２８で測光してメモリするようにしても、専用のランプ監視センサを用いてランプの分光エネルギー分布を常に修正して用いるようにしてもよい。さらに、このＲＯＭ３８には、カラーネガフィルムの分光感度分布（図１１参照）および被写体の固有ベクトル（図１４参照）等が記憶されている。
【００５５】
露光量決定装置３２は、画像データメモリ３０の書込みおよび読み出しタイミングを制御するように、画像データメモリ３０に接続されるとともに、測光器２８を駆動するように接続されている。また、入出力ポート３４は、駆動回路４８を介してネガキャリア２１に接続されるとともに、駆動回路５０を介して調光フィルタ６０に、駆動回路５４を介してブラックシャッタ２４に、それぞれ接続されている。また、入出力ポート３４には、キーボード４４，検出器５２およびＣＲＴ４６が接続されている。
【００５６】
次に、露光量決定装置３２のＲＯＭ３８に予め記憶された露光量制御ルーチンを、図１５を参照して説明する。ネガキャリア２１にカラーネガフィルム２０が装填されてスタートスイッチがオンされると、ステップ１００において、駆動回路４８によってネガキャリア２１が駆動されることにより、カラーネガフィルム２０の搬送が行われる。カラーネガフィルム２０が搬送されている間に、検出器５２によってＤＸコードが読み取られるとともに、ノッチが検出される。
【００５７】
次のステップ１０２では、検出器５２によってノッチが検出されたか否かを判断し、ノッチが検出されたと判断されたときには、ステップ１０４においてカラーネガフィルム２０の搬送を停止することにより、カラーネガフィルム２０上のこま画像を焼き付け露光位置に停止させる。ステップ１０６では、測光器２８を駆動してネガフィルム２０の透過濃度を測光する。測光器２８は、３つのフィルタを備えているため、ネガフィルムの透過濃度が３つの波長帯に分光されて測光されることになる。
【００５８】
次のステップ１０８では、読み取られたＤＸコードによってフィルム種を判別し、判別されたフィルム種に対応するカラーネガフィルムの分光感度分布をＲＯＭから読み出し、測光器２８による測光値，推定した光源の分光エネルギー分布並びにカラーネガフィルムの分光感度分布を用いて、先に説明した評価値Ｖを求める。
【００５９】
すなわち、前述の式（３）により、ΔＥが最小となる加重係数ｂ_ｎ を求め、求めた加重係数ｂ_ｎ を用いて前述の式（４）により、分光反射率を復元する。この復元された分光反射率を用いて前述の式（５）から評価値Ｖを求める。なお、センサ出力値として実際に用いる測光器２８による測光値は、ネガフィルムの透過濃度値であるため、図１６に示すフィルムの濃度−露光量特性曲線を用いて露光量を演算し、分光反射率の復元に用いる。
【００６０】
次のステップ１１０では、先に説明したように、求めた評価値Ｖが最小となるような、光源の分光エネルギー分布を推定する。
【００６１】
次のステップ１１２では、ステップ１１０で推定された撮影光源の色温度に基づいて露光量を演算し、ステップ１１４においてこの露光量に基づいて、調光フィルタ６０を制御して露光量制御を行う。
【００６２】
ステップ１１４で、調光フィルタ６０による露光量制御が終了すると、ステップ１１６において全こまの焼き付けが終了したか否かを判断し、終了していないときはステップ１００に戻って上記ステップを繰り返し、終了している場合にはこのルーチンを終了する。
【００６３】
上記実施例において、測光器２８および画像データメモリ３０から構成される測光手段と、露光量決定装置３２とから装置を構成し、上述の露光量制御ルーチンにおいてステップ１１０までを実行させることによって、この装置は、撮影光源の分光エネルギー分布をデータとして出力する装置としても機能することになる。
【００６４】
また、上記実施例では、本発明を、自動プリンタに適用した場合を例として説明したが、上述の撮影光源の分光エネルギー分布推定装置を、写真カメラ等の撮影装置に備えてもよい。この場合、得られる撮影光源の分光エネルギー分布をフィルムに記録し、プリンタやディスプレイに表示する顕像装置等の複写装置において、フィルムから記録された分光エネルギー分布のデータを読み取るようにしてもよい。
【００６５】
【発明の効果】
以上説明したように本発明によれば、すなわち、分光エネルギー分布が未知の光源をも含めて、撮影に用いられた光源の分光エネルギー分布を推定して、実際の光源を推定することが可能になるという顕著な効果が得られる。
また、この光源エネルギー分布の推定方法により推定した光源のエネルギー分布を基に、複写感材に焼き付けるための最適な露光量を決定することを可能とする露光量決定方法を実現できるという実用的な効果も得られる。
【図面の簡単な説明】
【図１】本発明において用いる主成分ベクトルの例を示す図（その１）である。
【図２】本発明において用いる主成分ベクトルの例を示す図（その２）である。
【図３】本発明において用いる主成分ベクトルの例を示す図（その３）である。
【図４】本発明において用いる主成分ベクトルの例を示す図（その４）である。
【図５】本発明において用いる主成分ベクトルの例を示す図（その５）である。
【図６】実施形態における、推定した光源の分光エネルギー分布と実際の光源の測定値との比較図（その１）である。
【図７】実施形態における、推定した光源の分光エネルギー分布と実際の光源の測定値との比較図（その２）である。
【図８】実施形態における、推定した光源の分光エネルギー分布と実際の光源の測定値との比較図（その３）である。
【図９】実施形態における、推定した光源の分光エネルギー分布と実際の光源の測定値との比較図（その４）である。
【図１０】実施形態における、推定した光源の分光エネルギー分布と実際の光源の測定値との比較図（その５）である。
【図１１】カラーネガフィルムの分光感度分布の一例を示す図である。
【図１２】実施例に係る、光源の分光エネルギー分布推定結果を適用した自動プリンタの概略図である。
【図１３】カラーネガフィルムの固有ベクトルの分光分布の一例を示す図である。
【図１４】被写体の固有ベクトルの分光分布の一例を示す図である。
【図１５】一実施例の露光量制御ルーチンを示す流れ図である。
【図１６】カラーネガフィルムの濃度−露光量特性を示す図である。
【図１７】色温度と評価値Ｖとの関係を示す図である。
【符号の説明】
２０ フィルム
２６ ペーパ
２８ 測光器[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a light source energy distribution estimation method for determining a light source type that illuminates a subject and an exposure amount determination method using the same, and more specifically, illumination from information obtained by imaging a subject illuminated by the light source. The present invention relates to a light source energy distribution estimation method, a light source energy distribution estimation apparatus, and an exposure amount determination method.
[0002]
[Prior art]
The exposure amount when a photographic film (hereinafter simply referred to as film) image is printed on a copy-sensitive material such as photographic paper is determined by the amount of light received from the subject when the film is photographed, and is different for each frame. In order to obtain a print with good color reproducibility, it is necessary to correct the printing exposure amount according to the photographing conditions.
[0003]
For this reason, in general, the exposure amount when reproducing a color image from a color original image to a copy light-sensitive material is determined by using a photometric device equipped with a color separation filter composed of a dye filter and a vapor deposition filter. The gray balance is determined by measuring the integrated transmission density of G) and blue (B) light and determining each of the R, G, and B light.
[0004]
However, since the shooting light quality information may change depending on the color feria such as the background and development conditions, the light quality cannot be estimated accurately, and the color reproducibility deteriorates due to the change in the light quality of the subject illumination light. There are things to do. This is because it cannot be determined which position on the film is gray.
[0005]
The most effective way to detect gray on this film is to estimate the color temperature of the photographic light source. In this regard, for example, Patent Document 1 relating to the application of the present applicant proposes a method for estimating the spectral energy distribution of a photographing light source from a signal of an image recorded under predetermined photographing conditions.
[0006]
This method is intended to estimate the color temperature of the light source used when photographing the subject. Specifically, the method estimates the color temperature of the light source from the information of the photographed image. . In short, the principle is to estimate the color temperature of the light source from the balance of R (red), G (green), B (blue), and what balance R and B are based on G. It is.
[0007]
That is, since the information of the captured image is the result of multiplying the spectral energy distribution of the light source and the spectral reflectance distribution of the subject, here the spectral temperature of the subject is assumed assuming the color temperature of the light source. The process of estimating the reflectance is performed for all types of light sources.
[0008]
However, when such a waveform prediction is performed on actually captured image data, unreasonable data may be obtained depending on the assumed light source. In other words, an unusual value such as a reflectance exceeding 100% or a negative value may be obtained.
[0009]
When such an abnormal value appears, the process of adding it as a penalty is performed for all the pixels in the image (or some extracted pixels), so that the color temperature is plotted on the horizontal axis. Thus, when the above penalty values are plotted, the result is as shown in FIG.
[0010]
In this figure, the color temperature corresponding to the point with the lowest penalty value is the light source with the highest probability of being the light source actually used for shooting estimated here. It will be.
The above is the outline of the technique disclosed in Patent Document 1.
[0011]
[Patent Document 1]
Japanese Patent Laid-Open No. 8-122157
[Problems to be solved by the invention]
The present invention aims to further improve the above-described technique to improve the estimation accuracy, and more specifically, the spectral energy distribution that has been impossible in the past is not known (that is, An object of the present invention is to estimate an actual light source by estimating a spectral energy distribution of a light source used for photographing including a light source whose spectral energy distribution is unknown.
[0013]
Still another object of the present invention is to provide an exposure amount determination method capable of determining an optimum exposure amount for printing on a copy-sensitive material based on the energy distribution of the light source estimated by the above-described light source energy distribution estimation method. Is to realize.
[0014]
[Means for Solving the Problems]
In order to achieve the above object, a light source energy distribution estimation method according to the present invention includes a plurality of predetermined function linear energy combinations represented by a linear combination of light sources, a photometric system spectral sensitivity, and a plurality of predetermined functions. A reference value determined by a product sum or an integral value of products of spectral reflectance distributions represented by a linear combination is obtained, and at least a part of reflected light from a light source for spectral energy distribution estimation is measured as a signal, and the reference value The spectral reflectance distribution that minimizes the difference between the measured value and the measured value obtained for the measurement is determined for each type of linear combination of the light source energy distributions, and the maximum value of the determined spectral reflectance distribution exceeds 1.0. The sum of the components is obtained as an evaluation value, and the linear combination of the light source energy distributions corresponding to the minimum value of the evaluation values is used as the energy distribution of the light source to be estimated for the energy distribution.
[0015]
Here, each of the plurality of predetermined functions is a principal component vector obtained from a plurality of light source data. As the principal component vector obtained from the plurality of light source data, it is preferable to use at least the third principal component, and it is preferable to use the fifth principal component in order to obtain higher accuracy. The reference value is preferably obtained in advance and stored in the storage means.
[0016]
The light source energy distribution estimation apparatus according to the present invention is represented by a linear combination of a plurality of predetermined functions, a spectral energy distribution of the light source, a spectral sensitivity of the photometric system, and a plurality of predetermined functions. A storage means for storing a product value of a product of the spectral reflectance distributions or a reference value determined by an integral value; a measuring means for measuring at least a part of the reflected light from the light source of the spectral energy distribution estimation target; A spectral reflectance distribution calculating means for calculating a spectral reflectance distribution that minimizes a difference between the reference value and a measured value obtained by measurement for each kind of linear combination of light source energy distributions, and the obtained spectral reflectance. An evaluation value calculating means for calculating the sum of abnormal components having a maximum distribution value exceeding 1.0 as an evaluation value, and a light source energy distribution corresponding to the minimum value of the evaluation values calculated by the evaluation value calculating means Combine, characterized in that a estimating means for estimating an energy distribution of the energy distribution estimation target of the light source.
[0017]
Here, each of the plurality of predetermined functions is a principal component vector obtained from a plurality of light source data. As the principal component vector obtained from the plurality of light source data, it is preferable to use at least the third principal component, and it is preferable to use the fifth principal component in order to obtain higher accuracy.
[0018]
The exposure amount determination method according to the present invention includes information on spectral energy distribution of a light source estimated by the light source energy distribution estimation method as described above, and estimation of a photographing light source energy distribution photographed on a film under a predetermined photographing condition. Based on photometric data obtained by measuring at least a part of the target image, the image to be printed is printed on the copy-sensitive material so that the gray of the image to be printed becomes gray under the estimated spectral energy distribution of the light source. It is characterized by determining the exposure amount for this.
[0019]
[Action]
Next, the operation of the present invention will be described.
[0020]
Assuming that the spectral energy distribution of the photographing light source is P (λ), the spectral sensitivity distribution of the film used for photographing is S (λ), and the spectral reflectance distribution of the photographing subject is ρ (λ), the exposure amount data E is expressed by the following equation: It can be expressed by (1).
[Expression 1]

[0021]
However, since the spectral reflectance distribution of the subject differs depending on the photographed image at the time of photographing, it cannot be specified, and it was difficult to estimate the spectral energy distribution of the photographing light source from the image on the film. Is as described above.
[0022]
In contrast, in the method for estimating the spectral energy distribution of a light source according to the present invention, the spectral energy distribution of a light source having an unknown spectral energy distribution is calculated based on the principal component analysis results of various known light sources called standard light sources. A component vector is obtained, and when this principal component vector is v _j (λ), a spectral spectrum distribution of the light source is generated according to the following equation (2).
[0023]
[Expression 2]

[0024]
Here, the k pieces of weighting coefficients b _n variously changing the light source in the spectral energy distribution of P (lambda) is generated in accordance with the above equation (2). With respect to the spectral energy distribution P (λ) of the light source thus generated, an evaluation value V is obtained by the method disclosed in Patent Document 1 described above. For the obtained evaluation value V, a weighting coefficient b _n giving the smallest V value is obtained.
[0025]
Then, the spectral energy distribution of the light source is calculated according to the above formula (2) for the weighting coefficient b _n that gives the minimum V value, and the obtained light source spectral distribution is estimated to be the spectral distribution of the light source used for photographing. To do.
The above is the outline of the operation of the present invention. More details are as follows.
[0026]
DETAILED DESCRIPTION OF THE INVENTION
If there is sensor output value data (E _ij ⁰ ) when photographing a subject illuminated with light from a light source illuminated with a light source having an unknown spectral energy distribution, the spectrum of the light source when the subject is recorded Assuming that the energy distribution is obtained, as shown in the following equation (3), the data E _ij ⁰ and the above-described weighting coefficient b _n are changed in various ways and generated according to the above-described equation (2). A solution is a light source having a spectral energy distribution that minimizes a difference ΔE from the sensor output value E _ij ^T corresponding to the spectral energy distribution P (λ) of the light source.
[0027]
[Equation 3]

[0028]
In order to perform the ΔE minimization calculation of the above-described equation (3), here, as the model calculation, six light sources of standard light A, D65, C of JISZ8720 and light D50, D55, D75 of auxiliary standard are used. Principal component analysis was applied to a total of 18 types of light source spectral spectra with 12 types of fluorescent lamps described in JISZ8719, and the principal component vectors were obtained.
[0029]
The characteristics are shown in FIGS. Here, FIG. 1 shows the first principal component, FIG. 2 shows the second principal component, and so on up to the fifth principal component.
With respect to the principal component vector thus obtained, spectral spectrum distributions of various light sources are generated according to the above-described equation (2).
[0030]
Here, a large number of light source spectral distributions P (λ) are generated according to the above equation (2) by varying the k weighting factors b _j described above. With respect to the spectral energy distribution P (λ) of the light source thus generated, ΔE in the above equation (3) was minimized in the case where 24 colors were used as the subject. Here, the data E _ij ⁰ is obtained from the following equation (4) obtained by modifying all the equations (1) by using the spectral sensitivity of the actual film and the actually measured spectral reflectance.
[0031]
[Expression 4]

[0032]
Incidentally, in the various b _n value in the corresponding light source spectral energy distribution P of (lambda), if include those showing the abnormal value on the spectral reflectance of the subject to be restored as corresponding thereto There is. That is, the obtained spectral reflectance ρ _i (λ) may be a negative value or significantly exceed 1.0.
[0033]
Since the spectral reflectance ρ _i (λ) does not exceed a negative value or 1.0 (that is, 0 ≦ ρ _i (λ) ≦ 1), conversely, this abnormal spectral reflectance ρ _i (λ It can be assumed that the spectral energy distribution P (λ) of the light source that is the source of) deviates from the spectral energy distribution P (λ) of the “true” light source.
[0034]
For this reason, the evaluation value V shown by the following formula (5) is introduced.
[Equation 5]

[0035]
The spectral light sources this evaluation value V, calculated for various b _n spectral energy distribution P of the light source corresponding to the value of (lambda), corresponding to a combination of b _n of the evaluation value V is minimized The energy distribution P (λ) is estimated as the spectral energy distribution P (λ) of the light source used for actual photographing.
[0036]
6 to 9 show the spectral energy distribution of the light source obtained by the above-described method and the so-called “true” spectral energy of the light source used when the image on the target film was actually taken. The distribution is shown in contrast.
Here, the case where the above-described principal component vector is used up to the third principal component and the case where up to the fifth principal component are used are shown for comparison.
[0037]
First, FIG. 6 is obtained by applying the estimation method according to the present invention to a spectral energy distribution (shown by a solid line) of a general xenon lamp and an image on a film photographed using the spectral energy distribution. It is a figure which compares the spectral energy distribution (it shows with the dashed-dotted line) of the "estimated" light source. Here, the case where the principal component vector is used up to the third principal component is shown.
[0038]
FIG. 8 shows a case where the above-described combination is used up to the fifth principal component. As is clear from the comparison between the two, the actual use of the principal component vector up to the fifth principal component (FIG. 8) compared to the case where the principal component vector is used up to the third principal component (FIG. 6). There is a slight difference in the degree of coincidence between the characteristic and the estimated result, but even when the principal component vector is used up to the third principal component, it is practically sufficiently usable.
[0039]
Next, FIG. 7 and FIG. 9 show the case of photographing in a situation where both a commercially available fluorescent lamp and a tungsten bulb are used. In these figures, the solid lines indicate these two. The spectral energy distribution of the combined state in the state where both the light sources are lit, and the one-dot chain line indicate the “estimated” light source obtained by applying the estimation method according to the present invention in this situation. Spectral energy distribution.
[0040]
FIG. 7 shows a case where the principal component vector is used up to the third principal component, and FIG. 9 shows a case where the principal component vector is used up to the fifth principal component. Even in this case, when the principal component vector is used up to the fifth principal component (FIG. 9) compared to the case where the principal component vector is used up to the third principal component (FIG. 7), the actual characteristics and estimation are performed. Although there is a slight difference in the degree of coincidence with the result, even when the principal component vector is used up to the third principal component, it is practically usable.
[0041]
Next, for reference, the result of estimating the spectral energy distribution of the light source by the method according to the present invention using the principal component vector up to the second principal component under the conditions shown in FIGS. 7 and 9. Is shown in FIG.
As is clear from FIG. 10, the estimation method using the principal component vector up to the second principal component is insufficient in accuracy of estimation and is not practical.
[0042]
From the above results, in the method for estimating the spectral energy distribution of the light source according to the present invention, it is preferable to perform estimation using at least the principal component vector up to the third principal component, and more preferably, the principal component vector is the fifth principal component. It can be said that estimation is performed using components.
[0043]
As described above, the information on the spectral energy distribution of the light source obtained by the method of estimating the spectral energy distribution of the light source according to the present invention can be applied to the control of the exposure time of the print at the time of photographic printing (printing). Is possible.
[0044]
In addition, when printing an image to be printed on a print sensitive material as a copy sensitive material, assuming an application of controlling the exposure time of the print based on the light source having the estimated spectral energy distribution, The above-described sensor spectral sensitivity distribution S _j (λ) can be applied to a film sensitivity distribution, for example, a color negative film sensitivity distribution as shown in FIG.
[0045]
Here, since the spectral energy distribution of the light source for measuring the color density from the film and the spectral sensitivity distribution of the printed light-sensitive material are obtained by measuring in advance, if the spectral transmittance distribution of a different film depends on the film, The density measured by a photometric device having a spectral sensitivity distribution equal to the spectral sensitivity distribution of the printed photosensitive material can also be obtained. This technique is also described in a method for estimating the spectral distribution of a film described in JP-A-4-310942.
[0046]
【Example】
Hereinafter, specific examples will be described.
In this embodiment, the present invention is applied to an automatic printer.
[0047]
FIG. 12 is a schematic view of an automatic printer that prints a photograph by the above-described printing method based on the estimated spectral energy distribution of the light source.
A lamp house 10 having a mirror box 18 and a halogen lamp is arranged below the color negative film 20 loaded in the negative carrier 21 and conveyed to the printing unit. A dimming filter 60 is disposed between the mirror box 18 and the lamp house 10. As is well known, the dimming filter 60 is composed of three filters: a Y (yellow) filter, an M (magenta) filter, and a C (cyan) filter.
[0048]
Above the negative film 20, a lens 22, a black shutter 24, and a color paper 26 are arranged in this order. Light that has been irradiated from the lamp house 10 and transmitted through the light control filter 60, the mirror box 18, and the color negative film 20 An image is formed on the color paper 26 by the lens 22.
[0049]
On the side edge of the color negative film 20, a DX coat representing the type of the color negative film is recorded and a notch is formed. In order to detect the DX code and the notch, a detector 52 composed of a light emitting element and a light receiving element is arranged so as to sandwich the side edge of the negative film 20.
[0050]
A photometer 28 is arranged in a direction inclined with respect to the optical axis of the imaging optical system described above and at a position where the image density of the color negative film 20 can be measured. The photometer 28 includes three filters having a center wavelength of 450 ± 5 nm, 550 ± 5 nm, and 700 ± 5 nm and a half width of 15 to 50 nm, respectively, and a two-dimensional image sensor. With this photometer 28, the light transmitted from the color negative film can be spectroscopically measured in three wavelength bands.
[0051]
The photometer 28 is connected to an exposure amount determination device 32 constituted by a microcomputer via an image data memory 30 for storing image data measured by the photometer 28. The exposure amount determining device 32 includes an input / output port 34, a central processing unit (CPU) 36, a read only memory (ROM) 38, a random access memory (RAM) 40, and a data bus and a control bus for connecting them. A bus 42 is provided.
[0052]
The ROM 38 stores a program of an exposure amount control routine described below and the distribution of eigenvectors e ₁ (λ), e ₂ (λ), e ₃ (λ) of the color negative film shown in FIG. 13 for each film type. ing. Although FIG. 13 shows the distribution of eigenvectors for one type of film, the same applies to other types of films. The ROM 38 stores in advance the spectral energy distribution of the halogen lamp in the lamp house 10, the spectral sensitivity distribution of the paper used, and the three spectral sensitivity distributions of the photometer corresponding to the transmission wavelength ranges of the three filters. ing.
[0053]
When changing the paper, the spectral sensitivity distribution of a plurality of types of paper may be stored in advance in the ROM, and the spectral sensitivity distribution of the paper to be used may be selected using the keyboard. The spectral sensitivity distribution of the necessary paper stored in the memory may be read into the RAM. It is also preferable to change the spectral energy distribution of the lamp when the lamp is replaced.
[0054]
For this purpose, the spectral energy distribution of the lamp may be measured and memorized by the photometer 28 directly or through a filter, or the spectral energy distribution of the lamp may always be corrected and used by using a dedicated lamp monitoring sensor. Good. Further, the ROM 38 stores the spectral sensitivity distribution of the color negative film (see FIG. 11), the eigenvector of the subject (see FIG. 14), and the like.
[0055]
The exposure amount determining device 32 is connected to the image data memory 30 and to drive the photometer 28 so as to control the writing and reading timing of the image data memory 30. The input / output port 34 is connected to the negative carrier 21 via the drive circuit 48, connected to the dimming filter 60 via the drive circuit 50, and connected to the black shutter 24 via the drive circuit 54. Yes. A keyboard 44, detector 52 and CRT 46 are connected to the input / output port 34.
[0056]
Next, an exposure amount control routine stored in advance in the ROM 38 of the exposure amount determination device 32 will be described with reference to FIG. When the color negative film 20 is loaded on the negative carrier 21 and the start switch is turned on, the negative carrier 21 is driven by the drive circuit 48 in step 100, whereby the color negative film 20 is conveyed. While the color negative film 20 is being conveyed, the DX code is read by the detector 52 and a notch is detected.
[0057]
In the next step 102, it is determined whether or not a notch has been detected by the detector 52. If it is determined that the notch has been detected, the conveyance of the color negative film 20 is stopped in step 104, whereby the color on the color negative film 20 is stopped. The top image is stopped at the printing exposure position. In step 106, the photometer 28 is driven to measure the transmission density of the negative film 20. Since the photometer 28 has three filters, the transmission density of the negative film is spectrally divided into three wavelength bands and photometrically measured.
[0058]
In the next step 108, the film type is discriminated based on the read DX code, the spectral sensitivity distribution of the color negative film corresponding to the discriminated film type is read from the ROM, the photometric value obtained by the photometer 28, and the estimated spectral energy of the light source. The evaluation value V described above is obtained using the distribution and the spectral sensitivity distribution of the color negative film.
[0059]
That is, by the above equation (3), determine the weighting coefficients b _n where ΔE is minimized, using the weighting coefficients b _n determined by the foregoing equation (4), to recover the spectral reflectance. Using this restored spectral reflectance, the evaluation value V is obtained from the above equation (5). Since the photometric value obtained by the photometer 28 actually used as the sensor output value is the transmission density value of the negative film, the exposure amount is calculated using the density-exposure characteristic curve of the film shown in FIG. Used for rate restoration.
[0060]
In the next step 110, as described above, the spectral energy distribution of the light source is estimated so that the obtained evaluation value V is minimized.
[0061]
In the next step 112, an exposure amount is calculated based on the color temperature of the photographing light source estimated in step 110, and in step 114, the light control filter 60 is controlled based on this exposure amount to perform exposure amount control.
[0062]
When the exposure amount control by the dimming filter 60 is completed in step 114, it is determined in step 116 whether or not all the frames have been printed. If not, the process returns to step 100 to repeat the above steps. If so, this routine ends.
[0063]
In the above-described embodiment, the apparatus is constituted by the photometric means constituted by the photometer 28 and the image data memory 30 and the exposure amount determination device 32, and this step is executed by executing up to step 110 in the above-described exposure amount control routine. The device also functions as a device that outputs the spectral energy distribution of the imaging light source as data.
[0064]
In the above embodiment, the present invention is applied to an automatic printer as an example. However, the above-described spectral energy distribution estimation device for a photographing light source may be provided in a photographing device such as a photographic camera. In this case, the obtained spectral energy distribution of the photographic light source may be recorded on a film, and the spectral energy distribution data recorded from the film may be read in a copying apparatus such as a visual display device displayed on a printer or display.
[0065]
【The invention's effect】
As described above, according to the present invention, that is, it is possible to estimate an actual light source by estimating a spectral energy distribution of a light source used for photographing including a light source whose spectral energy distribution is unknown. A remarkable effect is obtained.
Further, it is possible to realize an exposure amount determination method capable of determining an optimum exposure amount for printing on a copy-sensitive material based on the light source energy distribution estimated by the light source energy distribution estimation method. An effect is also obtained.
[Brief description of the drawings]
FIG. 1 is a diagram (part 1) illustrating an example of a principal component vector used in the present invention.
FIG. 2 is a second diagram illustrating an example of principal component vectors used in the present invention.
FIG. 3 is a third diagram illustrating examples of principal component vectors used in the present invention.
FIG. 4 is a diagram (part 4) illustrating examples of principal component vectors used in the present invention;
FIG. 5 is a diagram (part 5) illustrating examples of principal component vectors used in the present invention;
FIG. 6 is a comparison diagram (part 1) between an estimated spectral energy distribution of a light source and an actual measured value of the light source in the embodiment.
FIG. 7 is a comparison diagram (part 2) between the estimated spectral energy distribution of the light source and the actual measured value of the light source in the embodiment.
FIG. 8 is a comparison diagram (part 3) between the estimated spectral energy distribution of the light source and the actual measured value of the light source in the embodiment.
FIG. 9 is a comparison diagram (part 4) between the estimated spectral energy distribution of the light source and the actual measured value of the light source in the embodiment.
FIG. 10 is a comparison diagram (part 5) between the estimated spectral energy distribution of the light source and the actual measured value of the light source in the embodiment.
FIG. 11 is a diagram illustrating an example of a spectral sensitivity distribution of a color negative film.
FIG. 12 is a schematic diagram of an automatic printer to which a spectral energy distribution estimation result of a light source is applied according to an embodiment.
FIG. 13 is a diagram illustrating an example of a spectral distribution of eigenvectors of a color negative film.
FIG. 14 is a diagram illustrating an example of a spectral distribution of an eigenvector of a subject.
FIG. 15 is a flowchart showing an exposure control routine according to an embodiment.
FIG. 16 is a diagram showing density-exposure amount characteristics of a color negative film.
17 is a diagram illustrating a relationship between a color temperature and an evaluation value V. FIG.
[Explanation of symbols]
20 Film 26 Paper 28 Photometer

Claims (8)

Product sum or integral of the product of spectral energy distribution of light source expressed by linear combination of a plurality of predetermined functions, spectral sensitivity of photometric system, and spectral reflectance distribution expressed by linear combination of a plurality of predetermined functions The standard value determined in
Measure at least a part of the reflected light from the light source for spectral energy distribution estimation,
Spectral reflectance distribution that minimizes the difference between the reference value and the measured value obtained by measurement is determined for each type of linear combination of the light source energy distribution,
The sum of abnormal components having a maximum value of the obtained spectral reflectance distribution exceeding 1.0 is obtained as an evaluation value,
A light source energy distribution estimation method, wherein a linear combination of light source energy distributions corresponding to a minimum value of the evaluation value is an energy distribution of a light source to be estimated for energy distribution. The light source energy distribution estimation method according to claim 1, wherein each of the plurality of predetermined functions is a principal component vector obtained from a plurality of light source data. The light source energy distribution estimation method according to claim 2, wherein at least a third principal component is used as the principal component vector obtained from the plurality of light source data. The light source energy distribution estimation method according to claim 1, wherein the reference value is obtained in advance and stored in a storage unit. Product sum or integral of the product of spectral energy distribution of light source expressed by linear combination of a plurality of predetermined functions, spectral sensitivity of photometric system, and spectral reflectance distribution expressed by linear combination of a plurality of predetermined functions Storage means for storing the reference value determined in
Measuring means for measuring at least part of the reflected light from the light source of the spectral energy distribution estimation target as a signal;
Spectral reflectance distribution computing means for computing the spectral reflectance distribution that minimizes the difference between the reference value and the measured value obtained by measurement for each type of linear combination of the light source energy distribution, and the obtained spectral reflectance distribution Evaluation value calculation means for calculating the sum of abnormal components having a maximum value exceeding 1.0 as an evaluation value;
Light source energy, comprising: an estimation means for estimating a linear combination of light source energy distributions corresponding to the minimum value of the evaluation values calculated by the evaluation value calculation means as the energy distribution of the light source to be estimated for the energy distribution. Distribution estimation device. 6. The light source energy distribution estimation apparatus according to claim 5, wherein each of the plurality of predetermined functions is a principal component vector obtained from a plurality of light source data. The light source energy distribution estimation apparatus according to claim 6, wherein at least the third principal component is used as the principal component vector obtained from the plurality of light source data. Information on the spectral energy distribution of the light source estimated by the light source energy distribution estimation method according to any one of claims 1 to 4,
A light source in which the gray of the image to be baked on the photographic film is estimated based on photometric data obtained by photometry of at least a part of the image of the photographic light source energy distribution estimation target photographed on the photographic film under predetermined photographing conditions A method for determining an exposure amount, comprising: determining an exposure amount for printing on a copy-sensitive material so as to be gray under the spectral energy distribution of

Images